Subplasmalemmal endoplasmic reticulum controls K_Ca channel activity upon stimulation with a moderate histamine concentration in a human umbilical vein endothelial cell line

Abstract

This study was designed to elucidate the role of the subplasmalemmal endoplasmic reticulum (sER) in autacoid-induced stimulation of Ca²⁺-dependent K⁺ channels in the umbilical vein endothelial cell-derived cell line EA.hy926. Cells were transfected with the Ca²⁺ probe cameleon targeted to the ER for visualization of the ER network. A patch pipette was then placed close to or far (> 5 μm away) from the sER, single channel recordings (patch clamp technique) were monitored simultaneously with measurements of either ER Ca²⁺ concentration (using the Ca²⁺ probe Cam4-ER) or cytosolic free Ca²⁺ concentration ([Ca²⁺]_i; using fura-2) using a deconvolution imaging device. A voltage-dependent, large conductance Ca²⁺-dependent K⁺ channel (BK_Ca; single channel conductance (γ), 250 pS) was found. At membrane potentials of +40 and −40 mV, the EC₅₀ for Ca²⁺ was 2.7 and 49.7 μm, respectively. In the vicinity of the sER, the BK_Ca channel activity induced by 10 μm histamine was 32 times higher (open probability (P_o) = 0.083 ± 0.026) than in areas away from the sER (P_o = 0.0026 ± 0.002). However, at supramaximal histamine stimulation (100 μm), BK_Ca channel activation was similar in patches in the vicinity of or away from the sER (P_o = 0.18 ± 0.09 and 0.25 ± 0.07, respectively). In contrast to BK_Ca channel activity, ER Ca²⁺ depletion (Cam4-ER) and elevation of [Ca²⁺]_i in response to 10 and 100 μm histamine were not influenced by the pipette position. We conclude that in endothelial cells, the activation of BK_Ca channels in response to moderate histamine concentration essentially depends on the proximity of the sER domains to the mouth of this K⁺ channel. These findings further support our concept of the subplasmalemmal Ca²⁺ control unit (SCCU) and add the local activation of Ca²⁺-activated K⁺-channels to the function of the SCCU.

In many non-excitable cells, such as endothelial cells, an elevation in cytosolic Ca²⁺ concentration is a pivotal event which leads to cell activation. In endothelial cells this Ca²⁺ elevation triggers the synthesis of relaxing (nitric oxide, prostacyclin, endothelium-derived hyperpolarizing factors) or contracting (endothelin) factors (Graier et al. 1994). As in most non-excitable cells, agonist stimulation of endothelial cells yields inositol 1,4,5-trisphosphate (IP₃)-mediated Ca²⁺ depletion of the endoplasmic reticulum (ER) that triggers activation of a Ca²⁺ influx pathway that is termed capacitative Ca²⁺ entry (CCE; Putney, 1986). The molecular nature of this non-voltage-gated Ca²⁺ entry, as well as a direct link between Ca²⁺ store depletion and activation of Ca²⁺-permeable channels in the plasma membrane is still under debate (for review see Putney et al. 2001). However, Ca²⁺-mediated negative feedback inhibition of CCE channels has frequently been described and it has been suggested that mitochondria buffer Ca²⁺ at the mouth of the CCE channel in order to prevent Ca²⁺-mediated inhibition (Rizzuto et al. 1993, 1994; Rutter et al. 1993; Lawrie et al. 1996; Gilabert & Parekh, 2000). While this concept is sufficient to explain how the Ca²⁺-inhibitable CCE channels might be insulated from rises in cytosolic Ca²⁺, one needs to consider that the amount of Ca²⁺ actually entering the cell via the CCE pathway critically depends on the driving force for Ca²⁺ (Lückhoff & Busse, 1990). Upon autacoid stimulation, the driving force for Ca²⁺ is provided by membrane hyperpolarization due to a stimulation of Ca²⁺-activated K⁺ channels (Lückhoff & Busse, 1990; Groschner et al. 1992). Thus, in order to allow sufficient Ca²⁺ entry into the endothelial cell, subplasmalemmal Ca²⁺ concentration needs to be buffered (by mitochondria?) to avoid Ca²⁺-mediated inhibition of CCE channels while at the same time increases in subplasmalemmal Ca²⁺ must initiate stimulation of Ca²⁺-activated K⁺ channels to provide membrane hyperpolarization. Consistent with these deductions, we have recently shown in the human umbilical vein-derived endothelial cell line EA.hy926 an insulated subplasmalemmal Ca²⁺ elevation that occurs in response to histamine (10 μm) (Paltauf-Doburzynska et al. 1998, 2000). Such localized Ca²⁺ elevation seems to contribute greatly to the activation of the iberiotoxin-sensitive large conductance Ca²⁺-dependent K⁺ current (BK_Ca) in response to moderate (i.e. 10 μm) but not supramaximal (i.e. 100 μm) histamine concentrations (Frieden & Graier, 2000). To account for the localized Ca²⁺ gradient restricted between the plasma membrane and the superficial domains of the ER we suggested the existence of a subplasmalemmal Ca²⁺ control unit (SCCU) (Graier et al. 1998).

In the present study we challenge the hypothesis that subplasmalemmal domains of the ER functionally communicate with the plasma membrane in order to activate Ca²⁺-activated K⁺ channels in EA.hy926 cells. Thus, we set out to determine whether the localization of BK_Ca channels on the cell surface and their proximity to the ER network is important for channel activation in response to different histamine concentrations. For this purpose, the endothelial ER network was visualized by transfecting cells with the Ca²⁺ probe cameleon targeted to the ER (Cam4-ER; Miyawaki et al. 1997). This allowed us to monitor Ca²⁺ depletion of the ER at the same time as recording single channel BK_Ca current (in the cell-attached mode) with the patch pipette positioned close to or far from the ER network. Using this approach we were able to demonstrate that activation of endothelial BK_Ca channels by moderate but not supramaximal histamine concentrations essentially depends on the existence of subplasmalemmal ER domains in the vicinity of the K⁺ channel and, thus, further supports the concept of localized ion regulation in endothelial cells.

METHODS

Materials

The human umbilical vein endothelial cell line EA.hy926 (Edgell et al. 1983) was a gift from Dr Cora-Jean S. Edgell, Pathology Department, University of North Carolina, Chapel Hill, NC, USA. Cell culture chemicals were obtained from Life Technologies, Vienna, Austria, and Fetal calf serum (FCS) was from PAA Laboratories, Linz, Austria. The acetoxymethyl ester form of fura-2 (fura-2 AM) and ER-Tracker Blue-White DPX were purchased from Molecular Probes Europe BV, Leiden, The Netherlands. Histamine was obtained from Sigma Chemicals, Vienna. Cam4-ER was a kind gift from Professor Roger Tsien (Howard Hughes Medical Institute, University of California at San Diego).

Cell culture and transfection

The human umbilical vein endothelial cell line EA.hy926 (Edgell et al. 1983) was used for this study. Endothelial cells of passage 45 and higher were grown on glass coverslips (24 mm diameter) in Dulbecco's minimum essential medium (DMEM) containing 10 % fetal calf serum, 4.5 mg l⁻¹ d-glucose and 1 % HAT (5 mm hypoxanthin, 20 μm aminopterin, 0.8 mm thymidine). After reaching 80 % confluence endothelial cells were transiently transfected with cDNA encoding Cam4-ER using SuperFect (Qiagen, BioTrade, Vienna, Austria) in the standard transfection protocol described for this transfection reagent.

Measurement of the Ca²⁺ concentration in ER and cytosol

Between 24 and 36 h after transfection cells were used for experiments. Prior to the experiments, the cells were washed twice with storage buffer (SB; DMEM containing 2 % horse serum and 0.1 % of a vitamin mixture (Life Technologies, Vienna, Austria) and 0.2 % essential amino acids). All experiments were performed at room temperature (20–23 °C).

Cytosolic Ca²⁺ concentration

For measuring the cytosolic free Ca²⁺ concentration ([Ca²⁺]_c), conventional fura-2 experiments were performed in single Cam4-ER transfected endothelial cells as described previously (Graier et al. 1998; Paltauf-Doburzynska et al. 1998, 1999). Briefly, cells were loaded for 45 min at room temperature in the dark in SB containing 2 μm fura-2 AM, washed twice and equilibrated for a further 30 min in SB. For further procedures the coverslip was mounted into an experimental chamber and perfused (1 ml min⁻¹) with Hepes-buffered solution (HBS) containing (mm): 145 NaCl, 5 KCl, 2.5 Ca₂Cl, 1 MgCl₂ and 10 Hepes acid, pH adjusted to 7.4.

ER Ca²⁺ concentration ([Ca²⁺]_er)

For monitoring the free Ca²⁺ concentration in the ER, Ca²⁺-dependent fluorescence resonance energy transfer (FRET) of the cameleon construct was monitored (Miyawaki et al. 1997; Arnaudeau et al. 2001). The coverslip was washed twice, transferred into the experimental chamber and perfused with HBS.

Data acquisition and analysis

Experiments were performed using a deconvolution microscope described previously (Graier et al. 1998; Paltauf-Doburzynska et al. 1998). Briefly, the microscope consists of a Nikon inverted microscope (Eclipse 300TE, Nikon, Vienna) equipped with CFI Plan Fluor × 40 oil immersion objective lens (NA 1.3, Nikon, Vienna, Austria), an epifluorescence system (150 W XBO; Optiquip, Highland Mills, NY, USA), a computer controlled z-stage (Ludl Electronic Products, Haawthrone, NY, USA) and a liquid-cooled CCD camera (−30 °C; Quantix KAF 1400G2, Roper Scientific, Acton, MA, USA). Image resolution was 0.171 μm pixel⁻¹. Excitation wavelengths were selected using a computer-controlled filter wheel (Ludl, Electronic Products, Hawthrone, NY, USA). All devices were controlled either by Metafluor 4.0 (Visitron Systems, Puchheim, Germany) or a customized version of IonWizard (IonOptix, Milton, MA, USA) software for Ca²⁺ measurements and ImagePro 3.0 software (Media Cybernetics, Silver Spring, MD, USA) for deconvolution imaging.

To avoid vibration of the filter wheel on the emission side when cameleon FRET was measured ratiometrically, which would harm the simultaneous measurement of ionic currents, no ratiometric measurements of the cameleon FRET were performed. In addition, due to interference between the excitation filter of Ca²⁺-free fura-2 (i.e. 380 nm) and the blue fluorescent protein of the cameleon construct, no ratiometric cytosolic Ca²⁺ measurements could be performed in cameleon transfected cells. Fluorescence dye bleaching was corrected by calculating the bleaching function for each individual dye in every single experiment using a one phase exponential equation for curve fitting. The degree of bleaching of the Cam4-ER signal was ∼2.1 % within the experimental time of 4 min. The cameleon Ca²⁺ measurements are represented as F/F₀, where F is the fluorescence at a given time and F₀ is the mean fluorescence of five individual measurements 10 s after acquisition has started. For fura-2 measurements, the F/F₀ was used to calculate the actual cytosolic free Ca²⁺ concentration ([Ca²⁺]_i) using a calibration curve obtained in situ in Cam4-ER-transfected endothelial cells.

To monitor [Ca²⁺]_c and [Ca²⁺]_er cells were illuminated at 340 (fura-2; 340HT15; Omega Optical, Brattleboro, VT, USA) and 440 nm (cameleon; 440AF21; Omega Optical), respectively. Emission was monitored at 510 (510WB40 with dichroic 430DCLP; Omega Optical) for fura-2 and 535 nm (535AF26 with dichroic 455DRVP; Omega Optical) for Cam4-ER (FRET). Quantitative measurements of fura-2 and Cam4-ER were not performed simultaneously, but in the case of fura-2 all cells used were transfected with Cam4-ER.

Image analysis and deconvolution were performed as previously described using Image-Pro 3.0 software (Media Cybernetics, Silver Spring, MD, USA) and the constrained iterative Tikhonov-Miller algorithm (Microtome, VayTek, Fairfield, IA, USA).

Patch clamp recording whilst ER/cytosolic Ca²⁺ imaging

We used the cell-attached and inside-out configurations of the patch clamp technique (Hamill et al. 1981). Borosilicate glass pipettes were pulled on a Narishige puller (Narishige Co Ltd, Tokyo, Japan), fire polished and had a resistance of 6–10 MΩ. Sealing of the cells was performed in HBS and currents were recorded with an EPC-7 amplifier (List Medical, Darmstadt, Germany) filtered at 1 kHz (900C9L8L, Frequency Devices, Haverhill, MA, USA), digitized by a digidata 1320 interface (Axon Instruments, Union City, CA, USA) and sampled by a PC running pCLAMP 8.0 software (Axon Instruments) at 5 kHz. Single current analysis was performed using Fetchan and pStat software (Axon Instruments). The channel open state probability (P_o) was expressed as the time spent in the open state (t_o) divided by the total time of the recording (t): P_o = t_o/t. P_o was usually calculated on a 2 s sweep. When several identical channels (N) were simultaneously open on the same patch, the open probability of one channel was calculated as: P_o = (t_o1 + 2t_o2 + 3t_o3 +… + Nt_oN)/Nt. Where t_oN is the time spent by a channel at the open level N. Pipette solution contained (mm): 130 KCl, 1 MgCl₂, 10 Hepes (adjusted to pH 7.4 with KOH). The standard experimental bath solution contained (mm): 130 KCl, 1 MgCl₂, 2 CaCl₂, 8 Hepes (adjusted to pH 7.45 with NaOH). To modify the K⁺ equilibrium potential the following bath solutions were used (mm): 130 NaCl, 5.6 KCl, 1 MgCl₂, 2 CaCl₂, 8 Hepes (adjusted to pH 7.45 with NaOH). For determination of the Ca²⁺ dependency of the channel, solutions containing 500 nm to 30 μm free Ca²⁺ were prepared, using MaxChelator program (Stanford, CA, USA) for calculating the appropriate amount of EGTA and Ca²⁺ to add in standard bath solution.

Loading the endothelial cells with ER-Tracker Blue-White DPX

Cells were loaded with ER-Tracker Blue-White DPX according the manufacturer's standard procedure (Molecular Probes, Leiden, The Netherlands). Briefly, cells were grown on glass coverslips and loaded for 30 min at room temperature in the dark in SB containing 500 nm ER-Tracker Blue-White DPXe, washed twice and equilibrated for further 10 min in SB. For further analysis, cells were mounted into an experimental chamber and perfused with HBS as indicated above.

Statistics

Analysis of variance (ANOVA) was performed and statistical significance was evaluated using Scheffé's post hoc F test. Level of significance was defined as P < 0.05.

RESULTS

Single channel characterization of the K⁺ channel

Recently, ourselves and others have described the existence of large conductance Ca²⁺-activated K⁺ channels in Ea.hy926 cells (Haburcak et al. 1997; Kamouchi et al. 1999; Frieden & Graier, 2000). In this study, a single channel characterization was performed in the inside-out configuration. In symmetrical high K⁺ concentrations the conductance of the channel was 251.8 ± 8.5 pS (n = 11). Changing the K⁺ concentration of the bath solution from 130 mm K⁺ to 5.6 mm K⁺ shifted the reversal potential of the current from 0.6 to approximately +80 mV (extrapolated reversal potential) according to the Nernst equation (Fig. 1A).

Figure 1. Characterization of large conductance Ca²⁺-dependent K⁺ channel.

A, current-voltage relationship obtained from inside-out patches in high symmetrical K⁺ (130 mm; linear regression fitting, r² = 0.99), or in low intracellular K⁺ (5.6 mm; polynomial 2nd order fitting, r² = 0.99). Each point is the mean of 6–10 single experiments. B, open state probability (P_o) of the BK_Ca channel expressed as a function of holding membrane potential in the presence of various Ca²⁺ concentrations: 2 mm, 30 μm, 1 μm and 500 nm. Recordings were performed in high symmetrical KCl concentration. Each point shows the mean ± s.e.m. of 3–7 experiments. Curves were fitted according a Boltzmann equation for sigmoid curves (2 mm, 30 and 1 μm) or an exponential fit (500 nm) using Prism 3.0 (r² = 0.86–0.88; GraphPad 3.0). C, correlation between P_o and the Ca²⁺ concentration in the bath in an inside-out patch at holding potentials of +40 and −40 mV. Points represent means ± s.e.m.(n = 4–8). Curves were fitted according to the equation for sigmoid concentration-response curves using Prism 3.0 (GraphPad 3.0; r² = 0.98 and 0.94 for aV_h (Hp) of −40 and +40 mV, respectively).

To demonstrate that this type of K⁺ channel can indeed serve as a monitor for the subplasmalemmal Ca²⁺ concentration at the intracellular mouth of the channel, channel activity in the inside-out configuration was correlated to the Ca²⁺ concentration of the bath solution. The channel activity was increased by elevating the free Ca²⁺ concentration in the bath solution. Figure 1B shows the mean open probability of the channel as a function of the holding potential for different free Ca²⁺ concentrations. The EC₅₀ values for Ca²⁺ activation of the BK_Ca channel in EA.hy926 cells at holding potentials of +40 and −40 mV were 2.7 and 49.7 μm, respectively (Fig. 1C) and, thus, were in ranges similar to those described for endothelial BK_Ca channels (for review see Nilius et al. 1997). For estimation of subplasmalemmal Ca²⁺ concentration ([Ca²⁺]_sub), the following sigmoid concentration-response curve was used:

To test the contribution of the BK_Ca channel activity to the histamine-evoked membrane hyperpolarization of the endothelial cells, current clamp recordings in the absence and presence of iberiotoxin, a selective inhibitor of the BK_Ca channels, were performed. In these experiments, 10 μm histamine induced long-lasting membrane hyperpolarization of 21.5 ± 1.5 mV (from a resting membrane potential of −30.6 ± 3.4 to −52.1 ± 3.2 mV; n = 6) in the absence of iberiotoxin. In the presence of 100 nm iberiotoxin, the histamine (10 μm)-evoked membrane hyperpolarization was reduced by approximately 66 % to 7.2 ± 2.1 mV (from a resting membrane potential of −28.1 ± 2.8 mV to −35.3 ± 3.9 mV; n = 6; P < 0.05vs. in the absence of iberiotoxin).

Distribution of the cameleon probe to the endoplasmic reticulum

The endothelial cells were transiently transfected with the Ca²⁺ probe Cam4-ER (Miyawaki et al. 1997). This allowed us to visualize the architectural organization of the ER, as well as to measure changes in the ER Ca²⁺ concentration. Figure 2 confirms that the Ca²⁺ probe was indeed target the ER. A double staining with the ER fluorescent vital dye ER-Tracker Blue-White DPX (Grimaldi et al. 1999) shows a good match between the ER marker and Cam4-ER.

Figure 2. Subcellular localization of the Cam4-ER probe.

Localization of the Cam4-ER (left image, green) and the vital dye ER-Tracker Blue-White DPX (middle image, red) in cultured endothelial cells. Cells were transfected with cDNA encoding for the Cam4-ER probe and were loaded with 250 nm ER-Tracker Blue-White DPX for 30 min at 37 °C. Cam4-ER and ER-Tracker Blue-White DPX were monitored at 480 ± 30 and 380 ± 15 nm excitation and 535 ± 26 and 510 ± 40 nm emission. Images show shadow projections of selected slices from image stacks of 10 × 300 nm wide z sections which were deconvoluted with the iterative constrained Tikhonov-Miller algorithm using the theoretical point spread function. The overlay (right image) function of Image-Pro software (Media Cybernetics) was then performed.

Simultaneous measurement of ER localization, [Ca²⁺]_er and BK_Ca channel activity

Based on the visualization of the ER network the patch pipette was placed in the vicinity of central (i.e. perinuclear) or peripheral ER domains (Fig. 3A) or at central and peripheral locations where subplasmalemmal ER structures were at a distance of more than 5 μm away (Fig. 3B). The cells were stimulated with 10 μm histamine at a holding potential of +40 mV to enable monitoring of BK_Ca channel activity. In order to verify the presence of BK_Ca channels in each individual patch, the existence and/or the number of BK_Ca channels present under the patch pipette was verified in inside-out configuration after each stimulation. Seventy to seventy-five per cent of all the patches (close and far from the ER) contained BK_Ca channels and the number of BK_Ca channels per patch found in the vicinity of the sER (2.8 ± 0.4, n = 7) did not differ significantly from that obtained in patches far from sER domains (3.6 ± 0.5, n = 9).

Figure 3. Localization of the patch pipette with regard of the ER.

EA.hy926 cells were transiently transfected with the Ca²⁺ probe Cam4-ER. Two days after transfection cells were used for the experiments. Prior to establishing a seal, the architectural organization of the ER was analysed by visualization of Cam4-ER fluorescence at 440 ± 21 nm excitation and 535 ± 26 nm emission by performing a stack of five z sections (300 nm interslice distance) followed by image deconvolution. This procedure was performed within 5 min for each individual cell that was used for experiments. Thereafter, the pipette was placed to establish a seal on plasma membrane regions that were close to (A) or far from (B) ER domains. Left images show the Cam4-ER fluorescence and right images represent a phase contrast picture indicating the pipette location on this particular cell. After the seal has been established patch clamp recordings and imaging were performed as described in Methods.

If the pipette was located close to an sER structure the BK_Ca channel stimulation in response to histamine (Fig. 4A) was much more pronounced compared with experiments where the pipette was placed on a part of the membrane without visible sER domains (Fig. 4B). In particular, BK_Ca channel activity expressed as its maximal P_o was 0.083 ± 0.026 (n = 7; estimated [Ca²⁺]_sub ∼400 nm) in patches close to sER domains and 0.0026 ± 0.002 (n = 9; estimated [Ca²⁺]_sub ∼80 nm; P < 0.05vs. close to sER) when the pipette was placed in an ER-lacking area (Fig. 4C). There was no difference in the basal P_o close or far from visible sER domains (P_o = 0.001; estimated [Ca²⁺]_sub ∼70 nm). The analysis of the kinetics of BK_Ca channel activation in response to 10 μm histamine in the vicinity of the sER revealed a transient activation of K⁺ currents (Fig. 4D). Notably, there was no considerable difference between the kinetics and intensity of BK_Ca channel activation when the pipette was located in the vicinity of central (maximal P_o = 0.078 ± 0.031; estimated [Ca²⁺]_sub ∼370 nm, onset time until maximal P_o = 63.0 ± 23.4 s and total activation time = 121.3 ± 33.1 s; n = 4) or peripheral ER (maximal P_o = 0.088 ± 0.051; estimated [Ca²⁺]_sub ∼410 nm, onset time until maximal P_o = 51.3.0 ± 16.5 s and total activation time = 142.3 ± 38.9 s; n = 3; Fig. 4D).

Figure 4. Effects of a moderate histamine concentration (i.e. 10 μm) on BK_Ca channel activity in membrane patches located close to or far from the ER.

Original recordings of the unitary current in EA.hy926 cells in response to 10 μm histamine recorded in the cell-attached configuration using a pipette solution containing (mm): 130 KCl, 1 MgCl₂, 10 Hepes (adjusted to pH 7.4 with KOH). Extracellular bath solution contained (mm): 130 KCl, 1 MgCl₂, 2 CaCl₂, 8 Hepes (adjusted to pH 7.45 with NaOH). A, the pipette was placed on cell membrane areas in the vicinity of intracellular domains of the ER. Tracings show the activity of the BK_Ca channel prior to histamine application (a) and during the time of maximal stimulation by 10 μm histamine that occurred 0.5–0.7 min (b and c) after histamine stimulation. B, the pipette was placed on a region of the cell membrane that was at least 5 μm away from an ER domain. Channel activity is shown before application of histamine (a) and during the time of maximal stimulation by 10 μm histamine (b and c). No further activation was observed within 5 min. To test the presence and functionality of BK_Ca channels in the patch, an inside-out configuration was established after each individual experiment (d). Only those experiments where the BK_Ca channel was observed in the inside-out configuration were taken in consideration. Closed and open levels of the channel are indicated to the immediate right of each recording by ‘c’ and ‘o’, respectively. C, statistical evaluation of the effect of 10 μm histamine on the maximal activity of BK_Ca channel when the pipette was close to intracellular ER domains (close to sER), and when it was far from the ER (far from sER). Each bar represents the mean ± s.e.m. (n = 7 close to and 9 far from sER). *P < 0.05 vs. close to sER. D, representative time courses of BK_Ca channel activation in the proximity of central (perinuclear) and peripheral sER domains of the cell, as well as far from the sER.

To test whether or not these differences in BK_Ca channel activity observed in the various patch locations were due to changes in the Ca²⁺ depletion of the ER, the reduction of ER Ca²⁺ content was monitored simultaneously to BK_Ca channel recordings. As shown in Fig. 5, histamine-induced depletion of the ER was independent from the pipette location and no differences in the amount and the kinetics of ER Ca²⁺ depletion was found between cells with the pipette in either of the two positions.

Figure 5. Effects of pipette positioning on the ER Ca²⁺ depletion induced by histamine.

Simultaneously to the single channel recording (shown in Fig. 4) the Ca²⁺ content of the ER and its depletion in response to 10 μm histamine was measured by monitoring the FRET appearance of the ER-targeted Ca²⁺ probe Cam4-ER at 440 ± 21 nm excitation and 535 ± 26 nm emission. No significant difference was observed in the Ca²⁺ store depletion in response to histamine between various pipette positions on the cell surface. Data points represent means ± s.e.m. (n = 14 close to and 12 far from sER).

Another explanation of the observed differences between the BK_Ca channel activation in response to histamine would be that the placement of the pipette affects global endothelial Ca²⁺ signalling. To challenge this hypothesis, the global cytosolic Ca²⁺ concentration ([Ca²⁺]_i) was monitored in Cam4-ER-transfected cells loaded with fura-2. As shown in Fig. 6A, the fura-2 fluorescence could be monitored separately from the Cam4-ER fluorescence, which allowed [Ca²⁺]_i to be monitored while the pipette was located close to or far from the sER. Histamine-induced elevation of [Ca²⁺]_i was not affected by the pipette location ([Ca²⁺]_i increase in response to 10 μm histamine: pipette close to sER, 225 ± 35 nm, n = 7; pipette far from sER, 257 ± 26 nm, n = 7; Fig. 6B). In the absence of the pipette, the signal was identical ([Ca²⁺]_i increase in reponse to 10 μm histamine, 262 ± 51 nm, n = 7).

Figure 6. Effect of pipette positioning on the cytosolic Ca²⁺ signal induced by 10 μm histamine.

A, fluorescence images of fura-2-loaded endothelial cells transfected transiently with Cam4-ER. Right, fura-2 fluorescence visualized at 340 ± 15 nm excitation and 510 ± 40 nm emission. Middle, fluorescence image of the architectural organization of the ER of the cell indicated by Cam4-ER fluorescence. Left, overlay of the fura-2 (red) and Cam4-ER (green) fluorescences in this frame. Only the fura-2 signal of the transfected cell on which the patch pipette was placed was used. B, simultaneously to the single channel recording (shown in Fig. 4) the cytosolic Ca²⁺ concentration was measured by monitoring fura-2 fluorescence at 340 ± 15 nm excitation and 510 ± 40 nm emission. Endothelial cells were transiently transfected with cDNA encoding Cam4-ER. After 2 days in culture, cells were loaded with 2 μm fura-2 for 45 min at room temperature in the dark. In order to position the patch pipette with respect to the sER, the ER structure was visualized by monitoring the Cam4-ER fluorescence at 440 ± 21 nm excitation and 535 ± 26 nm emission. Image analysis was performed as described in the legend to Fig. 3. The effect of 10 μm histamine applied close to and far from intracellular ER domains on [Ca²⁺]_i was monitored by fura-2 fluorescence. Data points represent means ± s.e.m. (n = 7).

In contrast to our findings with 10 μm histamine, a supramaximal stimulation of the endothelial cells with 100 μm histamine achieved similar BK_Ca channel activation close to (Fig. 7A; P_o = 0.171 ± 0.091; estimated [Ca²⁺]_sub ∼900 nm) and far from (Fig. 7B; P_o = 0.251 ± 0.074; estimated [Ca²⁺]_sub ∼1300 nm) sER domains (statistics shown in Fig. 7C). The activation of BK_Ca channels by 100 μm histamine in the proximity of the sER (onset time until maximal P_o = 22.8 ± 7.62 s and total activation time = 137.2 ± 14.8 s; n = 6) was very similar to those distant from the sER (onset time until maximal P_o = 23.3 ± 6.0 s and total activation time = 132.1 ± 16.3 s; n = 8) showing similar transience and similar kinetics (Fig. 7D).

Figure 7. Effects of a supramaximal histamine concentration (i.e. 100 μm) on BK_Ca channel activity in membrane patches located close to and far from the sER.

A and B present original recordings in the cell-attached configuration of the unitary current in EA.hy926 cells in response to 100 μm histamine applied close to and far from the sER, respectively. Tracings show the activity of the BK_Ca channel before application of histamine (a) and during the time of maximal stimulation to 100 μm histamine that occurred 0.3–0.4 min (b and c) after histamine stimulation. C, statistical evaluation of the effect of 100 μm histamine on the maximal activity of BK_Ca channel when the pipette was close to intracellular ER domains (close to sER), and when it was far from the ER (far from sER). Each bar represents the mean ± s.e.m. (n = 6 close to and 8 far from sER). n.s., not significantly different vs. close to sER. D, representative time courses of BK_Ca channel activation close to and far from sER domains.

As seen with 10 μm histamine, neither global ER Ca²⁺ depletion (Fig. 8A) nor the global cytosolic Ca²⁺ signal ([Ca²⁺]_i increase in response to 100 μm histamine: pipette close to sER, 802 ± 195 nm, n = 3; pipette far from sER, 1007 ± 147 nm, n = 4; Fig. 8B) in response to 100 μm histamine were affected by the various pipette positionings.

Figure 8. Effect of pipette positioning on ER Ca²⁺ depletion and the cytosolic Ca²⁺ signal induced by 100 μm histamine.

A, simultaneously to the single channel recording shown in Fig. 7 the FRET appearance of Cam4-ER at 440 ± 21 nm excitation and 535 ± 26 nm emission was monitored to allow us to see the Ca²⁺ content of the ER and its depletion in response to 100 μm histamine. Data points represent the mean ± s.e.m. (n = 5 close to and 6 far from sER). B, simultaneously to the single channel recording shown in Fig. 7 the cytosolic Ca²⁺ concentration was measured with fura-2 (340 ± 15 nm excitation and 510 ± 40 nm emission). In transiently transfected endothelial cells the Cam4-ER fluorescence (440 ± 21 nm excitation and 535 ± 26 nm emission) was used to guide positioning of the patch pipette with respect to sER domains. As indicated cells were stimulated with 100 μm histamine.

DISCUSSION

In the present study the organelle-specific expression of a cameleon that senses ER Ca²⁺ concentration was combined with the patch clamp technique to elucidate the role of subplasmalemmal ER (sER) domains for activation of BK_Ca channels in endothelial cells. Our data strongly support our previous findings on localized subplasmalemmal Ca²⁺ gradients (Graier et al. 1998; Paltauf-Doburzynska et al. 1998) and indicate that the activity of the BK_Ca channel crucially depends on its proximity to sER domains. Thus we postulate, that subplasmalemmal Ca²⁺-containing organelles may constitute important regulators for transmembrane K⁺/ion currents.

In our previous work we have compared the EA.hy926 cell line with freshly isolated endothelial cells and found striking similarities in Ca²⁺ signalling, mechanisms of histamine-induced oscillations and endothelial nitric oxide synthase (eNOS) activation (Graier et al. 1998; Paltauf-Doburzynska et al. 1998, 2000). Additionally, EA.hy926 cells were found to express channels similar to those found in cultured human umbilical vein endothelial cells and also lack the β-subunit of the BK_ca channels as do many other isolated endothelial cells (for review see: Nilius & Droogmans, 2001). Thus, with the necessary caution, EA.hy926 cells may represent a reliable model to investigate endothelial Ca²⁺ signalling and Ca²⁺ function.

Recently, we have provided evidence that in endothelial and EA.hy926 cells an insulated subplasmalemmal Ca²⁺ elevation occurs in response to moderate agonist stimulation (i.e. 10 μm histamine; Graier et al. 1998). However, due to the limitations in the spatial resolution of the current techniques for measuring local Ca²⁺ gradients, direct visualization of endothelial spatial Ca²⁺ signalling is not yet possible. Thus, we have used whole-cell Ca²⁺-activated K⁺ currents to follow changes in the Ca²⁺ concentration in the whole subplasmalemmal space (Frieden & Graier, 2000). To allow spatial monitoring of subplasmalemmal Ca²⁺ concentration in well-defined subplasmalemmal regions of the cell, single endothelial BK_Ca channel recordings were utilized in combination with ER visualization. In order to use endothelial Ca²⁺-activated K⁺ channels as sensors of subplasamlemmal Ca²⁺ concentration, the K⁺ currents that are activated by histamine needed to be characterized. In inside-out configuration, we recorded a K⁺ channel with a conductance of ∼250 pS in symmetrical high K⁺ concentrations that corresponds to the BK_Ca class of channels (for review see Nilius et al. 1997). The Ca²⁺ sensitivity of the observed K⁺ channel is in the same range as those described for endothelial BK_Ca channels (for review see Nilius et al. 1997). The relatively low Ca²⁺ sensititvity of this BK_Ca channel is in agreement with recent findings (Papassotiriou et al. 2000) showing that in many endothelial cells (including EA.hy926 cells) BK_Ca channels do not contain the β-subunit, which leads to a reduced sensitivity of the channel for Ca²⁺ compared with BK_Ca channels present on other cell types. Overall, these data are in agreement with already published reports on BK_Ca channels in the Ea.hy926 cell line (Haburcak et al. 1997; Papassotiriou et al. 2000) and confirm our previous study in the whole-cell configuration where the existence of an iberiotoxin-sensitive K⁺ channel in the Ea.hy926 cell line was described (Frieden & Graier, 2000). The involvement of BK_Ca channels in membrane hyperpolarization of endothelial cells has already been described in endothelial cells from the rabbit aorta (Rusko et al. 1992), the intact rat aorta (Suh et al. 1999) and porcine coronary artery endothelial cells (Frieden et al. 1999). Furthermore, unlike the Ea.hy926 cell line where the BK_Ca channels seem to be the most prominent Ca²⁺-activated K⁺ channel, other types of K_Ca channel, such as SK_Ca and/or IK_Ca channels, were also shown to be involved in the change of membrane potential in other endothelial cell types (Groschner et al. 1992; Manabe et al. 1995; Marchenko & Sage, 1996). In line with the work of Nilius's laboratory on this cell line (Kamouchi et al. 1999), our data further indicate that the BK_Ca channels do not contribute to resting membrane potential while they represent the main current initiating membrane hyperpolarization in response to cell stimulation.

We have recently suggested that the subplasmalemmal Ca²⁺ gradient plays an important role in channel stimulation upon moderate cell stimulation, and thus extended the role of the SCCU as a regulator of K_Ca channels (Frieden & Graier, 2000). In the present study we went further and examined whether the location of the BK_Ca channel on the cell surface with respect to the intracellular ER network has an impact on channel stimulation. To achieve this aim, a deconvolution microscope was combined with conventional patch clamp instrumentation. In addition, cells were transfected with the fluorescent Ca²⁺ probe cameleon that was targeted to the ER (i.e. Cam4-ER; Miyawaki et al. 1997). Co-localization of the Cam4-ER fluorescence with that of the vital dye ER-Tracker Blue-White DPX (Grimaldi et al. 1999) revealed that in endothelial cells the Cam4-ER was exclusively targeted to the ER. This is in agreement with previous studies in other cell types which confirmed the accurate location of Cam4-ER in the ER (Miyawaki et al. 1997; Miyakawa et al. 1999; Arnaudeau et al. 2001). The expression of Cam4-ER in endothelial cells and the patch clamp/imaging system allowed us to visualize the ER network, to analyse the physical proximity of the patch pipette to the ER, to measure global ER Ca²⁺ depletion and to monitor single BK_Ca channels.

Remarkably, the activation of endothelial BK_Ca channels in response to moderate (i.e. 10 μm) but not supramaximal (i.e. 100 μm) histamine concentrations was much more pronounced in the proximity of the sER than at some distance from this Ca²⁺-containing organelle. This vast difference in the activation of the BK_Ca channels in response to 10 μm histamine might occur (i) because of an alteration in ER Ca²⁺ depletion and/or elevation in [Ca²⁺]_i due to the different pipette positioning or (ii) because for proper activation of BK_Ca channels a physical closeness to the sER domains is important.

One might expect that the position of the patch pipette with respect to the intracellular Ca²⁺ stores might affect ER Ca²⁺ depletion and/or elevation of [Ca²⁺]_i in response to autacoid stimulation. While such physical interference is possible and cannot be excluded entirely, our data strongly suggest that the location of the pipette does not alter histamine-induced intracellular Ca²⁺ signalling. Taking full advantage of our patch clamp-imaging system and of Cam4-ER-transfected endothelial cells, we could clearly show that despite differences in the local BK_Ca channel activation, the histamine-induced Ca²⁺ depletion of the ER remained unchanged at diverse pipette locations. Furthermore, histamine-evoked elevation of [Ca²⁺]_i, which was measured using the conventional fura-2 technique in Cam4-ER-transfected cells, remained unchanged despite the various locations of the patch pipette. Therefore it seems unlikely that differences in global cytosolic Ca²⁺ signalling or in Ca²⁺ depletion of the ER are responsible for the observed marked discrepancy in channel activation when the patch pipette was placed close to or far from sER domains.

Consequently, from the present data it seems that the proximity of sER domains to plasma membrane BK_Ca channels is key to achieving strong activation of these Ca²⁺-activated K⁺ channels under moderate stimulation. It appeared from our work that during modest endothelial cell stimulation (i.e. 10 μm histamine), there are distinct suplasmalemmal areas where Ca²⁺ increases up to an estimated 400 nm, that, in turn, leads to BK_Ca channel activation. On the other hand, during stimulation with 10 μm histamine, which raises the global cytosolic Ca²⁺ to ∼330 nm, the [Ca²⁺]_sub at a distance from sER domains remains < 100 nm free Ca²⁺ as indicated by the lack of BK_Ca channel activation in patches far from the sER. It seems likely that under moderate cell stimulation plasmalemmal Ca²⁺ pumps and/or the Na⁺-Ca²⁺ exchanger are able to prevent large subplasmalemmal Ca²⁺ elevation. In contrast, a supramaximal stimulation of the cells with 100 μm histamine which elevated cytosolic Ca²⁺ concentration to ≥ 1 μm initiated BK_Ca channel activation both close to and far from the sER. These data suggest that, at supramaximal stimulation of the cell the bulk Ca²⁺ elevation succeeded in triggering stimulation of Ca²⁺-activated plasma membrane channels regardless of the proximity to sER domains. Nevertheless, caution is necessary when comparing these Ca²⁺ concentrations, as the Ca²⁺ dependency of the BK_Ca channels was determined in the inside-out configuration, i.e. not in the ‘normal environment’ of the channel. Thus, further studies using newly developed methods for spatially monitoring the subplasmalemmal Ca²⁺ concentration (e.g. membrane-targeted cameleon) are necessary to further elucidate subplasmalemmal Ca²⁺ signalling.

There is evidence that certain plasma membrane ion channels/carriers can be distinctly distributed in the cell membrane in order to achieve a specific function (for review see Sheng & Pak, 2000). However, our present findings indicate, that in cultured EA.hy926 cells the BK_Ca channels are homogeneously distributed on the whole-cell surface. Indeed, in the inside-out configuration the patches contained equal numbers of BK_Ca channels no matter where the pipette was located initially. Moreover, neither the Ca²⁺ sensitivity nor the kinetics/activation of these channels differed at the different locations in the central versus peripheral regions of the cell. Thus, although other parameters like the surface/volume ratio or membrane composition may affect channel activity, the elevation of subplasmalemmal Ca²⁺ concentration constitutes the main regulator of BK_Ca channel activity in response to histamine stimulation.

These data further support our concept, that under moderate histamine stimulation (i.e. 10 μm) the BK_Ca channels become activated only in areas with vicinal superficial ER domains (i.e. SCCU). This observation may also be of particular importance in view of reports indicating that the channels responsible for CCE might be inhibited by elevated Ca²⁺ (for review see Berridge, 1995). Several reports suggest that mitochondria buffer subplasmalemmal Ca²⁺ elevation, thus, facilitating activity of the CCE channel (Hoth et al. 1997, 2000; Duchen, 1999, 2000; Gilabert et al. 2000). Notably, the amount of Ca²⁺ entering the endothelial cell via the CCE not only depends on the activity of CCE channels but also on the driving force for Ca²⁺ that is provided by membrane hyperpolarization due to the activation of Ca²⁺-activated K⁺ channels (Busse et al. 1988; Colden-Stanfield et al. 1990; Groschner et al. 1992, 1994). Thus, one might speculate that during moderate autacoid stimulation, the junction between sER domains and the plasma membrane (i.e. SCCU; Graier et al. 1998) may be important in providing the membrane hyperpolarization, while in other areas, possibly with mitochondria close to the membrane, Ca²⁺ entry through CCE channels might occur.

In conclusion, the present data clearly support our concept of localized subplasmalemmal Ca²⁺ elevation taking place in the endothelial cells (SCCU). By using a ER-targeted Ca²⁺ sensor (i.e. Cam4-ER) and simultaneous monitoring of single BK_Ca channels, it was possible to demonstrate a pivotal role for sER domains in the autacoid-initiated activation of BK_Ca channels in endothelial cells. Furthermore, our observation adds to the concept of CCE regulation and might explain, at least in part, the existence of subplasmalemmal areas of high and low [Ca²⁺] in one given cell at the same time.